ProjectSummary/Abstract Theproliferationofantibioticresistantpathogenshasincreasedattentiontotheuseofdrugcombinationstocombatthe evolutionofresistance.However,whilethephenotypicandgenotypicevolutionarypathstowardsmultidrugresistancecan be constrained by tradeoffs between resistance to one drug and susceptibility to other drugs or the humanhost environment, evolution often circumventstheseobstacles.FocusingonbothE.coliandclinicalspecies,wewillunravel the potential and limits of such approaches for constraining evolution. In Aim 1, using a novel selection device, the MEGAplate,whichfollowsmultiplediversifyingbacteriallineagesastheymigrateandevolveonlargeantibioticgradient landscapes, we will comprehensively map the repertoireofmultimutationalpathstohighlevelresistancetoarangeof antibiotics as well as to pairsofantibioticspresentingadaptivetradeoffs.Coupledwithwholegenomesequencingand automatedhighthroughputphenotyping,thisdevicewillenableustoidentifycommonandspecificadaptivemechanisms, revealthepredictabilityandfurtherevolutionarypotentialofeachmutationalstep,andtestwhetherchannelingevolution towards ?quasi deadend? genotypes can impede longterm adaptation. In Aim 2, we focus on cycling drug pairs presentingreciprocaladaptivetradeoffsandexaminevulnerabilitiesoftheapproach.Going?beyondtheaverage?,wewill construct a deep library of singlestep mutants selected on one of several different antibioticsandtestenmassetheir crossresistance to the other antibiotics. These results will identify rare ?escape? mutants which circumvent inherent tradeoffs between resistance to these drugs. Synthetically combining pairs of mutations which individually show resistance to one drug yet sensitivity to the other, we will reveal whether such mutations can interact nonadditively leadingtoresistancetobothdrugs.InAim3,weexaminetheeffectivenessyetweaknessesofnew?selectioninverting? compounds which we recently found to act preferentially against bacteria expressing the tetracycline resistance efflux pump. We will use a microfluidic device to follow single cells while switching between tetracycline and the new compounds,therebyidentifyingtheoptimalregimeforselectionagainstthepump.Wewillthensystematicallymutatethe tetracyclinepumptoidentifymutationsthatescapethetradeoff.Finally,inAim4,wefocusonthedifferencesbetweenin vitroandinvivoadaptivepathwaystoidentifytheevolutionaryconstraintsimposedbythehumanhostenvironment.We will use longitudinal isolates from clinical outbreakstoidentifywhichoftheresistantmutationsobservedinthelabalso appearduringpathogenevolutionwithinthehumanbodyandwhichothersareinvivoinaccessible.Comparingevolution of pathogens in immunocompetent versusimmunosuppressedpatientswillpointtoevolutionaryconstraintsimposedby innate immunity. In summary, our proposed research will reveal the genotypic and phenotypic constraints that govern evolutionary pathways towards multidrug resistance, both in the lab and during longterm infection in humans. Our longtermgoalistohelpdesigndrugregimesthatbetterpreventtheemergenceofresistanceandtodevelopalgorithms which based on the database of observedmutationalpathswillallowgenomebaseddiagnosticsofmicrobialinfections thatcanbothpredictcurrentresistanceprofileandanticipateitsfutureevolution.